Comparison of Some Engineering Properties of Expanded Polystyrene with Those of Soils

Transcription

1 TRANSPORTATION RESEARCH RECORD Comparison of Some Engineering Properties of Expanded Polystyrene with Those of Soils D. NEGUSSEY AND M. }AHANANDISH The engineering behavior of expanded polystyrene.(eps) wa investigated for potential applications as an alternative geomaterial. Backgrond on Eropean experience with EPS in road constmction over the pa t 2 years is provided. Tests were performed on EPS amples of two densities in constrained and nconfined deformation with loads applied in stress-controlled mode. Strength and deformation behavior and lateral tress coefficients for soils and EPS are compared. The reslts indicate that the engineering propertie of EPS can be qantified in a manner similar to those of earth materials. For some applications that involve infrastrctre rehabilitation and constrction of transportation facilities, EPS offers niqe advantages over oils. EPS needed for sbsrface con trction may contain recycled portions, and this wold be an important environmentaj incentive for sing EPS a a gcomaterial. Expanded polystyrene (EPS) is a synthetic material that is widely sed to manfactre disposable tensils and for prodct packaging. Most applications of EPS involve a short service life and one-time se with virtally no recycling. EPS is not a readily biodegradable waste prodct. There appears to be a good deal of interest in redcing EPS solid waste. EPS has pecliar characteristics that wold be desirable for sbsrface constrction applications. It is very light compared with soil and concrete and has energy absorption and inslation properties. EPS has been sed as sperlightweight fill and for fondation in lation to redce frost cover reqirements. Reported applications of EPS in sbsrface constrction have mostly been related to roads. There have been some large and small npblished applications in the United States. Norway has been the pioneer in EPS applications, and mch of the experience to date has been gided more by rle of thmb and field observation. Sorlie et al. (1) presented the Norwegian road constrction experience with lightweight soil sbstitte materials. They sggest that EPS shold have a compressive strength of 1 kpa at 5 percent deformation and an air resistance nmber of better than 7 to limit moistre pickp. Inflammability, dissolving by petrolem flids, and increased icing potential at near-freezing temperatres are practical problems that reqire design consideration. A fire-resisting variety of EPS can be specified at an additional cost of 5 to 1 percent above Department of Civil and Environmental Engineering, Syracse University, Syracse, N.Y standard qality (2). Potential damage of EPS by spilled fel can be mitigated by providing a membrane cover in addition to and below concrete slabs that are normally placed on top ofeps fills for improved load distribtion (3). king problem can be minimized by sing a thicker pavement strctre and by restricting moistre acces (J). A 4.5-m EPS fill is reported to have been sed as lightweight fill on soft grond for a temporary overpas bridge in Norway (4). Creep deformations at stres levels of p to 6 percent of yield were negligible. Transient heavy wheel load did not indce residal stresses. Sati factory performance of EPS as fill above rigid pipes to promote indced t(ench condition has been reported (5). This application is of special intere t becase EPS replaces the more commonly ed organic materials, which degrade in time. Frthermore, EPS for trench fill may consist entirely of sorted or recycled waste material. Rygg and SorHe (3) report on three application in Norway that involved repair o.f a road across a bog, new road constrction on a bog, and road embankment adjoining a bridge abtment, all sing EPS. These case histories indicate sccessfl se of EPS for road rehabilitation and constrction over difficlt fondation soils. EPS fills that remain sbmerged for extended drations are fond to retain abot 4 percent water by volme in the fir t year and abot 9 percent in 9 to 12 years (6). Even in cases of significant grondwater lowering, this level of moistre retention does not compromi e the sperlightweight advantages of EPS. However, adeqate cover mst be provided to prevent EPS breakot in times of sbmergence de to flooding or general rise in grondwater level. Norwegian experiences with EPS road fills over the past 2 year indicate aging effects to be insignificant within the de ign life of transportati.on facilities. The performance of road bilt on EPS fill is reported to have been satisfactory (7). Ftre trends envisioned for EPS in transportation inclde high embankments with steep ide slopes for concrete fonn work, floating bridges, and fill for bried strctres (8). EPS has also been sed for slope stabilization along troblesome transportation corridors in montaino contry. The option o.f sing recycled EPS or EPS with recycled traction in sbsrface constrction is a new concept that wold be worth exploring. Familiarity with EPS behavior and comp?ris?n wi~h geomaterials shold help promote broader applications m transportation and other geotechnical constrction.

2 44 TRANSPORTATION RESEARCH RECORD 1418 TEST MATERIALS EPS Prodction of EP bloc.i< begin from EPS pellets that contain a blowing agent. The pecific gravity, G., of polystyrene in amorphos or cry talline form i abot 1.1 (9). Th pellet arc fir t sbjected to team to fnrm prepffs. Initial pellet size and drntion of steam exposre determine the prepff and final block density. The prepff to pellet volme rntio is in the range of 4 to l. Approximately 1 percent of recycled EPS reclaimed from plant waste is hredded and mixed with the prepffs. The hredded EPS and prepffs are pored into a Teflon-lined molding box. Steam is injected throgh mall perforations along the moldjng box inside bondarie to indce additional expansion and fsion of the prepff. EP blocks of typically 1.25 x.6 x 4 m or 1.25 x.6 x 8 m are formed by this proce-..ommercially available den ities range from 15 to 5 kg/m 3, with 2 and 3 kg/m 3 varieties heing more common. Depending on qantity and location, the more common 2 and 3 kg/m 3 densities may cost between $25/m a:nd $5/m. 3, with the higher-density EPS costing more. Cylindrical EPS blocks 76 mm in diameter and 15 mm in height were provided for the inve tigation by Thermal Foams/ Syracse, Inc. Retaining Ring Base Drainage Poros Stone FIGURE 1 Oedometer with lateral pressre sensors. thin section fits inside the oedometer. With the spa1.:er block in place and restrained vertically, the instrmented ring can be pressrized. Tilt! 1esponsc of the strain g~ges 'N!!S calibrated against precise change in pre re et by a deadweight tester. The instrmented ection provide a sen irivity of abot 1 Mv/volt. The otpt is linear to lateral pressres in excess of 7 kpa. Clay A oft it1organic clay of medim plasticity was te ted Lo compare with the EPS behavior. The clay is sed for making pottery and has a water content of 3 percent and liqid and plastic limits of 4) and 24 percent, re pectively. The specijic gravity was determined to be 2.8 and the portion finer than 2µ. amonted to 46 percent by weight. The dry density of the clay sample wa abot 152 kg/111 3 Sand Reslts of tests on silica sand are al o compared with the behavior of EPS. The sand has a 5 of.55 mm and is niform, having a niformity coefficient C, of 1.9. Qartz is the predominant mineral in silica sand and individal grain are mo tly banglar. Tbe fine content of silica sand is le than 1 percent and the specific gravity is 2.6. The dry density of U1e. and sample was abot 155 kg/m 3 PROCEDURE EPS test samples 38 mm high and 76 mm in diameter were ct from EPS rods of the same diameter sing a hot nichrome wire. An EPS rod was placed in a glass tbe of a slightly larger diameter and having smooth ct perpendiclar ends. With the block held in position and the glass end as gide, test samples were ct. The weight and dimensions of ct samples were recorded. Samples for confined tests were installed in the oedometer, and those for nconfined tests were placed directly in the loading frame. In both cases vertical loading was applied pnematically in steps, and the applied load was sensed by a load cell positioned above the top cap. Vertical movement was monitored by a displacement transdcer. Confined compression tests were also performed on the clay and sand soils in the instrmented oedometer. The consolidation test on the clay soil was performed nder constant rate of displacement. The clay soil was also tested in nconfined compression. Procedres followed for the latter tests were in accordance with ASTM standards. EQUIPMENT Constrained deformation tests were performed sing a laterally instrmented oedometer 76 mm in diameter and 38 mm in height. Details of the oedometer are shown in Figre 1. Along the otside of a thin ection near the midheight of the oedometer strain gages are monted to sense lateral strain de to bending and hoop stres es. The in trmented ring adapt to a base fitted with a central poros stone and an oter -ring seal. A retaining ring clamps the ring to the base to seal. Drainage connection is provided along the base block. A spacer with a top -ring eal and narrow section along the RESULTS Test reslts from one-dimensional compres ion of a low-density EPS (21. kg/m 3 ) in Figre 2 show a behavior very similar to consolidation of clay soils. Segment of reloading, loading, and nloading are evident. A yield stres at an apparent maximm past pressre of abot 8 kpa can be identified. Before yield, void ratio tate changes are moderate a wold be in an overconsolidated stress range. Postyield, void ratio tate traverse in a path mch imilar to virgin compression and associated deformations are relatively large. Unloading and reloading are associated with smaller rates of void ratio change.

3 Negssey and Jahanandish ci :g 3 a::: EPS (21. kg/m') 1 1 Vertical Stress (kpa) FIGURE 2 Characlcri lie behavior of EPS nder one-dimensional deformation. The state of maximm past pressre is recalled on reloading, and the material shows evidence of having a "memory." One-dimensional compression test reslts for EPS amples of 21. and 3.4 kg/m 3 are compared in Figre 3. The pper crve represents a portion of the reslts shown in Figre 2 for the lower-density EPS. The lower crve repre ents reslt for the higher-density EPS and the yield sttess is abot 155 kpa. Neither sample bad a prior history of loading. The observed dijferences in apparent maximm pa t pres res are de to differenl prodct initial densities. Jn the manfactre of EPS blocks, densities are controlled by dration of expansion time rather than applied p.ressre. Yet initial density effects are mch like prestre s effects in soils. Constrained deformation test reslts for EPS are compared with those for silica sand and normally consolidated clay soil in the familiar semilog space [see Figre 4(a)]. Becase the void ratio contrast between EPS and soil is very large, the comparison is based on strain rather than void ratio. The reslts indicate that EPS behavior compares favorably with the behavior of soils in the stress range before yield. For loading past yield, the EPS compression is mch more severe even when compared with that of the clay. When the constrained deformation of EPS and the soils is compared in vertical stress and strain space (see Figre 4(b)J, the soil re -pon e crve show a tendency to stiffen with strain. As expected lhe sand develops mch less strain and stiffens more rapidly than the clay. The EPS maintains a relatively constant modls p to yielding near 155 kpa and almost 4 percent strain. Initial modli for the sand, clay, and EPS are approximately 25, kpa, 3, kpa, and 4, kpa, respectively. Being on the higher side the E P modls compares favoiably with the clay. TJ1e postyield EPS re ponse is characterized by a mch lower modls than the preyield. Even thogh confined, the EPS respon e crve simlates a behavior typical of nconfined compression. A plot of restraining lateral stress against applied vertical stress to the soils and the EPS is shown in Figre 5. Lateral tresses are highest in the clay and are lea t in the EPS. K vales for the clay, sand, and EPS are abot.55,.43 and.15 re pectively. l.n terms of cstomary approximations of K with reference to friction angle, the observed vales for the soils are in a general range common for normally consolidated clay and loose sand, respectively. However, the EPS response p to yield approximates a low Poisson's ratio material and closer to very tiff to hard clay or very dense sand. With frther loadjng beyond yield lateral tre ses in the EPS (and hence K states) decrea e a wold be th case for a 2 ~ 4 c ~ U1 6 t 8 1 A A o Vertical Stress (kpa) 1 1 a a a a a Sand ( 155 k<j/rqj) o o o o o Clay ( 152 kg/ ft1 j,.... EPS (3.4 kg/m ) (a) :g 3 a::: a 2 ~------\... EPS (21. kg/m 3l,..,...,... EPS (3.4 kg/m 3 3!2 "' ~ "' U1 81 :e (b) 8 a a R f 8 A 6.fl. ooo A ti. aooo cccao Sand A o o o o o o Cloy EPS (3.4 kg/m 3 ) 1 1 Vertical Stress (kpa) FIGURE 3 One-dimensional deformation behavior of EPS at two densities. Strain (jg) FIGURE 4 Comparison of one-dimensional deformation behavior of EPS with that of oils.

4 46 TRANSPORTATION RESEARCH RECORD ! 1 en en... "' VJ t' a a a o a Sand Cloy.,. EPS (3.4 kg/m') "' r:po o o O o o a a o o o a oo o co 5 1 Vertical Stress (kpa) n 15 FIGURE S Indced lateral stresses nder constrained deformation in EPS and soils. 2 material possessing a negative Poisson's ratio. This implies that lateral restraint is not essential to maintain teatres of constrained deformation in EPS and lends spport to the observation that constrained and nconfined deformation responses of EPS are alike [see Figre 4(b)J. Test reslts of constrained deformation and nconfined compression of EPS (3.4 kg/m 3 ) are compared in Figre 6( a) in natral vertical strain again t applied verlical stress space. These reslts are also presented in the cstomary semilog pace of void ratio and vertica I stress in Figre 6(b). The yield state in nconfined compres ion corresponds closely to the apparent maximm past pressre in the constrained test. As implied by the earlier interpretation of a low Poisson's ratio for EPS, constrained modls and Yong's modls are abot the same. Loading, nloading, and reloading responses are also similar for constrained and nconstrained conditions. Virgin compression in constrained mode resembles postyield nconfined loading response. A preferred shear plane does not develop, and lateral spreading does not occr in response to nconfined loading. The nconfined compre sion trength ofeps increa es with density (FigUl e 7). Compared with the soft clay nconfined compres ion strength, the low-density (21. kg/m 3 ) EPS is mch stronger and wold be eqivalent to a firm clay. The nconfined compression strength of the denser EPS (3.4 kg/ m 3 ) is closer to that of tiff to hard clay. There is generally small addi.tional increase in trength between 5 and 1 percent strain, bt lhi may depend on strain rate and load dration. EPS compression strength is sally reported at 5 percent strain. In practice, applied loads mst be distribted with a concrete pad or soil cover to prevent stress concentrations and pnching fa ilre lo realize the indicated high strengths. Figre S(a) shows nconfined compression reslts in which EPS contines to spport a stepwise increasing load past yield and at large strains. Postyield deformations are nonrecoverable and time dependent. Hysteresis loops are relatively small both at stales below and past yield [Figre 6(a)]. Ths damping and energy absorption featres of BPS are most favorable with virgin loading. As noted earlier, there was no evidence of a shear plane or lateral spreading at any stage of loading or deformation. The initial and final diameter of the sample was abot the same even thogh very large vertical strains occrred. 4 ~3 CL 6 en ~ "' 2 VJ ~ "'.25 :;:; :: }12 15 (a) EPS ( 3.4 kg/m 3 ),...,..,, Confined Unconfined Natral Vertical Strain (511) (b) EPS ( 3.4 kg/m' )..,...,, Confined Unconfined 1 1 Vertical Stress (kpo) FIGURE 6 Comparison of EPS compression behavior nder constrained and nconstrained lateral deformation. Reslts in Figre B(b) aj o show that EPS deformations have time dependence and that the degree of dependence is a fnction ofstre s level, tress history and void ratio. These reslts are for nconfined compression, bt the imilari ty of confined and nconfi ned compression behavior of EPS wa shown in Figre 6(a) and 6(b). Each load increment was CL 25 2 ~15 ""'" E c "' i7; "' 1 :::i 5 kg//m~) kg m) - 1'i:r;:;...,.,~...,.,~... ~~ T-rTT! 12 Ve rtical Strain (511) FIGURE 7 Comparison of nconfined compression strengths of EPS and clay. 8

5 Negssey and Jahanandish , l" ti 8 5 t :;:; Q:'. - 2 c; 1 (a) -- EPS (2 1. kg/m 3 ) Q+.-~~~~~~~~~~~~~~~~~~ - - EPS (21. kg/ m 3 ) (b) 1 1 Ve rtical Stress (kpo) (c) 25 tablished. The magnitde of creep deformations diminish with decreasing void ratio states. Isolated observations of time-dependent deformations at three selected stresses are presented in Figre 8(c) for EPS (21 kg/m 3 ). The selected stresses are at abot 5 percent of yield, yield at 83 kpa, and in postyield at 17 kpa. Elapsed times are referenced to load application. The crves resemble conventional primary and secondary creep stages for all three load steps. A tertiary type of creep stage may be inferred for the load steps corresponding to yield and postyield bt not for the 5 percent of yield or lowest load stage. For the first load step, the reslting strain over the entire load dration is reasonably small. Essentially all of the deformation in this stage occrs in the first few seconds and dring primary creep. Deformations in the secondary creep stage are relatively insignificant, and the stress level of 4 kpa is in the range of practical interest for geotechnical applications. A higher EPS density has been shown to reslt in a higher yield stress, and the corresponding stress threshold for time-dependent deformations shold be higher. Some of the observed engineering properties of EPS offer benefits sperior to soils for certain sbsrface applications. Fire hazard and chemical attack possibilities will reqire carefl consideration and attention to storage arrangements and constrction practice. Delivery of EPS to project sites and handling dring constrction will be easy. This shold make EPS se more attractive in projects where site access is difficlt. Environmental concerns and anticipated increased engineering activity in infrastrctre rehabilitation and development over the coming years will offer challenges and opportnities to adapt new constrction materials sch as EPS. In a manner similar to geotextiles and geomembranes, geofoams (EPS for geotechnical applications) may be the next wave of cost-effective and environmentally compatible constrction materials. 47 CONCLUSIONS ~ '=.;:. 2 (/) ~ 1 1 EPS (21. kg/m 3 ) 4 kp a a a a a a 8 3 kpo (Yield ) 17 kpo oooo DDDO~ o o o o o oooooooqc:.qj: 1 "' Time (sec) 1 FIGURE 8 Time-dependent behavior of EPS. sstained ntil time-dependent deformations sbsided. States of stress near and postyield show evidence of time dependence in that deformations occr withot change in stress state. Time-dependent deformations are Splall in the stress range below yield and along nloading and reloading segments when nloading is initiated before arresting time-dependent deformations; in sbseqent loading, time-dependent deformations commence when the previos maximm stress level is rees- 1. EPS is a sperlightweight material with a wide range of possible applications in geotechnical and infrastrctre engineering. 2. The stress-strain response and yield of EPS are well conditioned and bear close resemblance to some aspects of soil behavior. 3. Yong's modls and yield of EPS compare favorably with those of natral soils in the stress and strain range of practical interest for most geotechnical applications. 4. The yield stress for EPS is a fnction of initial density indced by the manfactring process and can be specified. 5. EPS behaves as a low Poisson's ratio material and has similar confined and nconfined compression response. 6. Dring confined compression, indced lateral stresses are very low, and hence EPS has a low K property. 7. Time-dependent deformations of EPS assme significance at stress levels approaching yield and postyield bt are reasonably small at working stress levels of abot half of yield. 8. EPS for sbsrface constrction may be derived, in part, from recycling. 9. EPS has been sed in Erope for geotechnical applications over the past 2 years, and the experience to date has been positive.

6 48 ACKNOWLEDGMENTS The athors thank Syracse University for providing fnds and facilities that enabled this investigation and Thermal Foams/ Syracse, Inc., for spplying EPS samples. The comments and sggestions for improvement made by the technical reviewers are acknowledged with gratitde. REFERENCES 1. Sorlie, A., R. G. Dahlberg, G. Refsdal, and. E. Rd. National Report: Norway. 16th World Road Congress, Vienna, Astria, Flaate, K. Sper Light Material in Heavy Constrction. Geotechnical News, Vol. 22, 1987, pp Rygg, N.., and A. Sorlie. Polystyrene Foam for Lightweight Road Embankment. 1th International Conference on Soil Mechanics and Fondation Engineering, Stockholm, Sweden, Vol. 2, 1lJ81, pp. '24'/-'lS2. 4. Skggedal, H., and R. Aaboe. Temporary Overpass Bridge Fonded on Expanded Polvstvrene. 1th Erovean Conference on Soil Mechani~s and Fondation Engineering, Florence; Italy, Vol. 2, 1991, pp Vaslestad, J. Load Redction on Bried Rigid Pipes. 1th Eropean Conference on Soil Mechanics and Fondation Engineering, Florence, Italy, Vol. 2, 1991, pp Van Dorp, T. Expanded Polystyrene Foam as Light Fill and Fondation Material in Road Strctres. International Congress on Expanded Polystyrene, Milan, Italy, Aaboe, R. Plastic Foam in Road Embankments: Experience with Expanded Polystyrene as a Light Weight Fill Material in Road Embankments. Norwegian Road Research Laboratory, Oslo, Norway, Refsdal, G. Plastic Foam in Road Embankments: Ftre Trends for EPS Use. Norwegian Road Research Laboratory, Oslo, Norway, Rdd, J. F. Physical Constants of Polystyrene. In Polymer Handbook (J. Brandrp and E. H. Immergt, eds.), 3rd edition, John Wiley, New York, DISCUSSION JOHNS. HORVATH Civil Engineering Department, Manhattan College, Bronx, N. Y I am actively involved in researching the application of rigid plastic foams to a wide variety of geotechnical problems. Sch materials are now recognized as geosynthetics nder the newly created prodct category of "geofoams." An inventory of geofoam materials and fnctions identified to date is smmarized elsewhere (1). EPS has been and still is the most widely sed geofoam (2,3). Among the reasons are its relative cost, environmental effects related to manfactre, finished prodct size, and material properties. A smmary of basic EPS material properties of interest to geotechnical engineers is presented elsewhere (4). In the light of this ongoing research, I wold like to comment on or qestion several items in this paper: 1. The most significant comment is a general one concerning the soil samples sed for comparison with EPS behavior. It is well established that the load-deformation be- TRANSPORTATION RESEARCH RECORD 1418 havior of soil is highly dependent on stress history, aging, relative density (for sand), and other factors in addition to particle size distribtion. Information concerning specimen preparation and conditioning was lacking for the soils tested. Conseqently, the conclsions drawn concerning whether EPS is behaviorally "better" or "worse" than soil are misleading. The conclsions of the paper in this regard are strictly applicable to two specific types of soil, each at a stress state that is ndefined. 2. Whereas recycling is an admirable goal, the contribtion of EPS to the domestic solid waste stream in the United States shold be kept in perspective. Recently pblished work detailing the scientific exhmation and evalation of actal landfills in the United States has demonstrated that all polystyrene prodcts combined occpy less than.25 percent of landfill volmes. In comparison, paper occpies approximately 4 percent. Frthermore, the potential for sing postconsmer recycled EPS for geofoam is complicated by the fact that in the United States, flame-retardant expandable polystyrene beads are sed rotinely to prodce EPS for constrctivtt prodcts sch as geofam (this is not tre id 5ViYie other contries, Norway being one example) bt normal beads for other prodcts. If flame-retardant EPS is commingled with normal EPS, the flame retardancy of the end prodct is compromised. Becase the two types of EPS are visally indistingishable, separation of the postconsmer recycled EPS is difficlt nless special measres (e.g., coloring) were to be implemented indstrywide dring manfactre. As noted in the paper, recycling of in-plant scrap prodced dring manfactring is already practiced by the EPS indstry. 3. On the basis of recent (1992) correspondence with both the Norwegian Road Research Laboratory and the world's largest manfactrer of expandable polystyrene beads for EPS (BASF), the air-resistance test is no longer sed for EPS geofoam qality control becase it was fond to provide inconsistent reslts. 4. Final, trimmed dimensions of EPS blocks in the United States are sally 61 by 129 by 2438 mm (2 by 4 by 8 ft). Molding lengths are 4.9 m (16 ft) or 7.3 m (24 ft). 5. Crrent definitions of EPS compressive strength or yield strength are based on tests performed at a relatively rapid strain rate. Indstry practice at present is to define EPS compressive strength as the stress at 1 percent strain, not 5 percent (5). The 5 percent strain criterion is sed in Norway and perhaps elsewhere. However, it is correct that the difference in compressive stress between 5 percent strain and 1 percent strain in the typical short-term test is relatively small. This is becase the elastic range ends and yielding begins for EPS at a compressive strain between 1 and 2 percent depending on prodct density. As will be discssed later, the compressive strength defined sing sch short-term tests does not provide insight into behavior nder the more typical geotechnical loads of long dration. 6. Was the inside of the instrment oedometer lined with any low-friction material or sbstance? Axial strains of EPS specimens in excess of 5 percent were implied for some of the one-dimensional compression tests, and friction along the wall of the oedometer, which wold redce the actal axial compressive stress on the test specimen, is of concern. 7. It was stated that constant-rate-of-strain loading rather than the traditional incremental loading was sed for the one-

7 Negssey and Jahanandish 49 dimensional compression tests on the clay specimen. What was the strain rate? 8. Incremental loading was sed for both the one-dimensional and nconfined compression tests on EPS specimens. What was the dration of each load increment? As noted previosly, estimation of yield stress for EPS is highly dependent on the load dration. 9. With reference to Figres 8(a) and 8(b), what were the drations of each load increment for which creep effects were observed? What criteria were sed to select a time for which the strain rate appeared to be zero? EPS is a thermoplastic material, and time-dependent deformations will contine for some as-yet-nknown dration at all stress levels. The qestion becomes whether sch time-dependent deformations are acceptable in a given application. Tests of EPS creep behavior in nconfined compression indicate that test dration of 1, hr or more are reqired to draw correct inferences a to whelber tertiary creep effects will occr at a given stress level (5). By comparison load dration ' of only abot O.l hr are shown in Figre 8(c). Figre 9 show the creep behavior for an EPS pecimen with a density close to that ed for the resl ts shown in the athor 'Figre 8(c). The "yield trengtb' of the pecimen in Figre 9 in the standard short-term test (strain rate of 1 percent/min) was slightly greater than 1 kpa. For stresses greater than abot 5 percent of yield, the long-term creep wold likely be excessive for many engineering applications. 1. The qalitative similarity between EPS and soil with regard to the shape of the load-deformation crve is not a niqe aspect of EPS. Rather, EPS and soil both exhibit rather classical material behavior. For example, as discssed elsewhere (6), annealed copper wire has behavior identical to that shown in Figre 2. The key difference is that for solids sch as EPS and copper wire, yield stress is bilt in dring manfactre, so soil mechanics concepts sch as "maximm past stress" have no physical relevance. On the other hand, for a particlate material sch as soil, yield stress is not niqe to the particlar soil bt depends primarily on stress history and other factors sch as aging. In smmary, I believe that the athors have provided some potentially sefl information concerning the relative behavior of EPS in nconfined verss one-dimensional axial compression if additional information concerning the type and dration of test loading is given. On the other hand, the comparisons with soil behavior are misleading becase of the lack of key soil mechanics information concerning stress history and so forth of the soil specimens. Conclsions as to the relative strength and stiffness of EPS and soil are probably,-..., Q. ~..._,,., Cl 1. - : "short-term" test 9 :' (approximate). ' '' 8 ' / load dration (hors): 7., 6... iii 5., 'iii... a. 4 E 3 :' reference: BASF (npblished data) Compressive Strain (%) 5 FIGURE 9 Creep behavior of 23.S-kg/m 3 EPS in nconfined axial compression.

8 5 impossible to generalize becase of the lack of niqe behavior for a given type of soil. REFERENCES 1. Horvath, J. S. Geofoam Geosynthetics: An Overview of the Past and Ftre. Oeosy111he1ics Worltl (in pres ). 2. Horvath, J. $.New Development in Geosynthctics: "Lite" Prodcts Come of Age. Stm1dardizmio11 News, ASTM, Sept. 1992, pp Horvath, J. S. Dark, No Sgar: A Well-Known Material Enters the Geo ynthetic Mainstream. Geotechnical Fabrics Report, IFAI, Oct. 1992, pp Horvath, J. S. Expanded Polystyrene (EPS) Geofoam: An Introdction to Material Behavior. Geote.rtiles and Geomembnmes (sbmitted for pblication). 5. Styropor; Process/11g- Measrements!Tests. TI -22 e. BASF AG, Ldwigshafen, Germany, Wood, D. M. Soil Behavior and Critical State Soil Mechanics. Cambridge University Press, Cambridge, United Kingdom, 199. AUTHORS' CLOSURE We thank the discssant for his interest in the paper and contribtion to a discssion. At this stage and in this context, the comparison of EPS behavior with that of soils is general. The two broad bt important classes of soils compared, clay and sand, are meant to be referenced generically. The range in compared behavior for a clean medim qartz sand, loose or dense, and a soft normally consolidated inactive clay wold not overlap nder conceivable sets of practical circmstances at corresponding stress levels. Physical, chemical, age, and stress attribtes have individal and collective inflence on the response and performance of a given soil. Drainage conditions, stress path, and loading rate wold also be important. In a strict sense, qantitative reslts for a soil apply to the soil in qestion and prevailing test conditions. That is why specific soils for specific applications are tested in practice, and this approach will no dobt contine, when jstified, in important applications that involve EPS and soils. However, sefl qalitative and, in some cases, practically adeqate qantitative observations are often drawn for a class of soils from reslts on specific soils throgh se of index properties. This is, of corse, common in geotechnical engineering practice and is the sense we attempted to convey in the comparison of EPS with a clay and sand soils. The argments made by the discsser against recycling EPS are nfortnate and represent a special interest position. At present, the spport for and awareness of the need for and benefits of recycling nonbiodegradable wastes are broad and do not reqire or frther jstification. For many sbsrface applications of EPS, flame retardancy wold not be critical. Potential compromise in flame retardancy de to mixing when recycling, as arged by the discssant, shold not be a reason for discoraging EPS recycling. The reference to air-resistance nmber made in the paper is in the context of a literatre srvey of pblished and acknowledged information. The discssant's information is presented as ndocmented personal commnication with sorces that are not well defined. Mention of a traceable sorce reference wold have been more sefl than the name TRANSPORTATION RESEARCH RECORD 1418 and rather flattering size description of a very large mltinational manfactrer. The behavior of EPS shows time dependence in preyield, at yield, and in postyield. This is shown by or reslts as well as by the npblished secondhand data frnished by the discssant. We have noted that creep effects beyond 5 percent of yield and in postyield may be potentially excessive for most geotechnical applications. The specification of yield for EPS will no dobt benefit from frther refinements in experimental observations. In testing soils, special techniqes are sed to mitigate the inflence of ndesirable conditions that violate test assmptions. Side friction redction in one-dimensional deformation tests by lbrication with vacm grease is a well-known procedre. A very important finding shown in the paper is the observation that lateral stresses are low before yield, compared with vertical stresses, and actally diminish in postyield. One-dimensional and nconfined compression behavior are shown to be similar, and lateral stresses at large strains and large vertical stresses become insignificant. The interface friction between EPS and a smooth metal srface is low in addition to lateral stresses being small and becoming negligible. In view of the foregoing, the discssant's expressed concern regarding adverse effects of side friction and sggested need for lbrication is, in or view, not rational. The strain rate for the constant rate of strain loading test on the clay soil was.6 percent/min. Load steps were maintained for 5 and 2 min dring incremental loading and nloading of EPS, respectively. Creep observations were made within the load increment time base. Longer tests and other areas of EPS behavior are the focs of present and ftre investigations. Uniqeness of engineering behavior is relative and contextal. The density, correspondence between constrained and nconfined compressions, lateral stress coefficients that develop dring confined compression, and other properties of EPS wold be considered niqe compared with soils. To forge a favorable connection between these niqe and important featres of EPS and soils, the senior athor proposed to representatives of the Society of the Plastics Indstry and others, at a workshop held at Syracse University in 1991, that EPS be referred to as geofoam. We were carefl to describe the observed breaks in onedimensional response of EPS as apparent maximm past pressre. Becase of the demonstrated correspondence between confined and nconfined behavior of EPS, yield and apparent maximm past pressre wold be synonymos since they apply to EPS only. The discssant views bilt-in stress in manfactred solids, and here he lmps copper and EPS, as not being analogos to maximm past pressre in soils. His reflections on this point are interestingly referenced to a textbook on soil behavior and critical state soil mechanics. The irony in the discssant's argment is that the development of critical state soil mechanics appealed to experimental stdies of Cam-clay, a manfactred clay with bilt-in stress history. Needless to say, the critical state soil model is alternatively known as the Camclay model. Pblication of this paper sponsored by Committee on Soil and Rock Properties.